Breast cancer pre- and post-treatment gene expresion anlyses

Introduction

Circulating Exosomal microRNAs as Predictive Biomarkers of Neoadjuvant Chemotherapy Response in Breast Cancer

`Neoadjuvant chemotherapy (NACT)`	an increasingly used approach for treatment of breast cancer
`Circulating Exosomal microRNAs(miRNA)`	small RNA molecules circulating in bodily fluids, potentially serving as biomarkers for treatment response after NACT
`Pathological Complete Response (pCR)`	absence of active cancer cells in a tissue from the tumor site after NACT
`Non-Pathological Complete Response (non-pCR)`	presence of active cancer cells in a tissue from the tumor site after the treatment

Introduction

Our Goals of the Study:

to compare gene expression and gene interaction networks for pre- and post- treatment
to conduct differential expression analysis across pCR and non-pCR groups

The Goal of the Paper:

to investigate whether circulating exosomal microRNAs could predict pCR in breast cancer patients treated with NACT

Materials and methods: Flowchart

Methods: Metadata & Describe

Methods: Describe

Methods: Gene Expression Analysis

Estimation of Size Factors
Estimation of Dispersion
Fitting the Negative Binomial GLM
Statistical Testing for Differential Expression
Multiple Testing Correction

Methods: Virtual pull-down and Gene Enrichement Analysis

Virtual pull-down:

library(DiscoNet)
network_pCR <- virtual_pulldown(
  seed_nodes = seed, 
  database = string_database, 
  id_type = "hgnc", 
  string_confidence_score = 700) # default value

Community detection

library(DiscoNet)
communities_list <- community_detection(
  graph_object, 
  algorithm = "mcode", # specifies algorithm, in this case MCODE
  D = 0.05, # threshold of weight percentage of the vertex to be use by MCODE standard value
  haircut = TRUE) # removes the nodes that are only connected to a single node in the community detected

Over representation analysis

library(fgsea)

ora_result_nonpCR <- fora(
    pathways = biological_process_list, 
    genes = gense_in_each_community, 
    universe = all_genes_in_analysis, 
    minSize = 10
)

Results: Gene Expression Analysis

Results: Virtual pull-down

Pre vs. Post Treatment:
- The nonpCR Group has 5 potential complexes or networks with upregulated or downregulated activity when comparing pre- and post-treatment
- The pCR Group has 11 potential complexes or networks with upregulated or downregulated activity when comparing pre- and post-treatment
Example plots:

Results: Gene Enrichement Analysis

For each relevant community in pCR and non_pCR the main enriched biological processes were identified
Representative plots:

Discussion and Conclusion

Differential gene expression was seen in higher levels in the pCR group
Further analysis into the communities found by enrichment analysis
Early diagnosis could help patients by:
- If pCR is achieved: perhaps surgery is not needed
- If pCR is not achieved: perhaps a different treatment method